Flox Intelligence Opens a $10 Million Series A for Edge AI That Keeps Moose Off the Tracks

The US Federal Aviation Administration estimates that wildlife strikes cost American civil aviation an average of $272 million a year between 1990 and 2025. In France, rail operator SNCF says wandering animals affect more than 8,000 trains annually and rank among its five leading causes of delays. A Stockholm startup called Flox Intelligence, presenting this week at TechCrunch Disrupt, is opening a $10 million Series A on the premise that edge AI can shrink those numbers.

The hardware is deliberately simple. A battery-operated unit sits dormant until a motion sensor wakes it. A camera captures the scene, on-device models identify which animal it is looking at, and the device plays back a species-appropriate sound intended to move that animal along. The responses vary over time, which the company says prevents animals from habituating to a repeated signal.

Founder Sara Nozkova spent thirteen years in transportation and autonomous vehicles before joining the AI lab at KTH Royal Institute of Technology, from which the company spun out in 2020. Her argument is that perception systems trained for engineered environments fail the moment they leave them. "We have taught machines to be brilliant, but only inside the environments we engineered for them," she told TechCrunch. "Stepping outside into nature I saw firsthand how the same machines quickly go blind."

About 150 units are now live across North America, Europe and Australia, at airfields and airport perimeters, along road and rail corridors, at fence ends and wildlife crossings, and in national parks, farmland, cities, mines and electrical substations. That is a modest footprint, and it is the number the Series A is meant to change.

The business model is hardware plus recurring software. Customers buy edge units up front and pay an annual subscription for the software and AI platform after the first year; the platform maps deployments, logs activity and generates the regulatory and biodiversity reports operators increasingly need. Flox also licenses to original equipment manufacturers, and train maker Alstom is running the software on trains, with field tests along Swedish railways that the manufacturer says showed potential to identify animals accurately and reduce collision risk.

The company forecasts roughly $1.5 million of revenue for 2026, a small figure against a $10 million raise, and it has raised about $5 million to date across pre-seed and seed rounds led by Unconventional Ventures and Tilia Impact Ventures, with GENIUS NY, Norrsken, Satgana and angels participating. The new round is intended to take deployments from roughly 150 devices to several thousand.

Flox is not alone in the niche. Robin Radar's Scarecrow plays species-specific distress calls and records bird activity, and Australia's iCHASE pairs computer vision with a moving laser that adjusts its response. Nozkova argues her moat is the breadth of the system and the field data behind it, which she calls ground truth on what actually makes a moose or a grizzly bear move safely. That claim rests on the company's own results; independent habituation data has not been published.

The investment case, as the company's own backers would acknowledge, is not about whether an AI detector is novel. It is about whether verified collision reductions, contract renewals and larger fleet orders follow the first 150 units. Nozkova's framing of where this goes is more ambitious than wildlife: "Physical AI is walking out of the factory and into the messy, alive outdoors," she said. "Almost nobody is building it." Proving that outdoors is harder than indoors is exactly what the next round of capital is for.